Child labor in the European periphery, 1870-2000

The “traditional view” of the history of child labor holds that children were rescued from child labor following the industrial revolution by activists and/or by the passage of effective child labor (minimum working age) laws. Since then, academics have suggested the importance of alternative key driving factors such as schooling, demography, household income or technology. Broadly, availability (and quality) of schooling is assumed to be inversely correlated to child labor practice and an increase in schooling is expected to translate into reductions in child labor. Demography, fertility and the share of children in the population changes the role of children in society by regulating child abundance and influences child labor practice. Lack of income may force families to send children to work until a minimum income threshold is reached, while the level and type of technology constrains the production system and eventually the contribution of children.

In this study, we aim at highlighting the trends but also the diversity of historical trajectories of child labor. For example, Humphries (2003) compares the evolution of child labor levels across GDP per capita and suggests that the historical English experience presented higher levels of child labor than the experience of current-day developing countries. Nevertheless, the narrow sampling of the studies on the historical trajectories of child labor is somewhat unsatisfying.

We expand the current literature by comparing the evolution of child labor in the understudied European periphery. The selected countries, Italy and Portugal, are the only (to our knowledge) European peripheral countries with available longitudinal child labor data (Goulart and Bedi, Social Science History, forthcoming). The comparison is useful because, a priori, wealthier Italy provides an upper limit of what happened in the European periphery, while Portugal provides a lower limit.

We use the method from Broadberry and Crafts (Economic History Review, 2003) to graphically explore the evolution of these countries against England, the GDP per capita leading country at the time. The most striking from this comparison with the estimates for Italy is the much tardier decline of child labor in Portugal regarding GDP per capita level. Independent of wealth, Portuguese children seem to have continued working much longer. To understand why, we use regression analysis with time series indicators for each of the five main drivers of the long-run evolution of child labor (legislation, education, demography, technology and income) to determine which driver was most important and when.